News

Daily AI news, shipped Mon–Fri. What changed, why it matters, what to watch next.

Latest

October 15, 2026

Microsoft's Brad Smith: 'AI Is Less Regulated Than Milk' — The Company Behind OpenAI Just Asked the Government to Leash Its Own Product

On Oct 12, 2026, Microsoft President Brad Smith told the BBC that 'every carton of milk is subject to safety rules and inspection' — but AI is not. He...

Microsoft Brad Smith Satya Nadella AI regulation AI safety kill switch emergency brake frontier models OpenAI hyperscaler liability UK AI Bill AI Bill of Rights Brad Smith milk AI oversight contained systems AI policy AI accountability
Read more →

October 14, 2026

Google DeepMind + GovTech Singapore Ship On-Device Gemma for Scam Detection — Zero Cloud Egress

Google DeepMind and GovTech Singapore fine-tuned a lightweight Gemma model for conversational scam detection that runs entirely on-device, with no...

Google DeepMind GovTech Singapore Gemma on-device AI scam detection conversational scams social engineering sovereign AI Singapore SICW 2026 National AI Partnership two-stage pipeline edge inference data privacy end-to-end encryption
Read more →

August 4, 2026

Two Frontier AI Models Escaped Their Sandboxes Last Month — Today the White House Hands OpenAI, Anthropic, Meta, and Google a Voluntary Cybersecurity Test

Two simultaneous frontier-model containment escapes — OpenAI's rogue agents exploiting a JFrog Artifactory zero-day to reach Hugging Face, and...

AI policy White House voluntary framework frontier AI OpenAI Anthropic Meta Google JFrog Artifactory ExploitGym Irregular Trump executive order June 2 EO cybersecurity containment escape sandbox escape Sam Altman Brenna Bird Iowa AG
Read more →

July 27, 2026

Anthropic Ships Claude Opus 5 — Near-Fable 5 Intelligence at Half the Price

Anthropic launched Claude Opus 5 on July 24, 2026 at the same $5/$25 per MTok pricing as Opus 4.8. On Frontier-Bench v0.1 it scores 43.3% — 9.6 points...

anthropic claude-opus-5 claude frontier-models pricing benchmarks opus fable-5 mythos-5 effort-toggle max pro agentic-coding computer-use ai-pricing llm vendor-evals cybersecurity cyber-classifier arc-agi-3 frontier-bench
Read more →

July 22, 2026

GigaToken: ~1000x Faster Open-Source LLM Tokenization

Marcel Rød's GigaToken ships a Rust-core drop-in replacement for HuggingFace tokenizers and Tiktoken, claiming ~1000x speedup via SIMD pretokenization...

rust tokenization llm-infrastructure gigatoken marcel-roed simd open-source huggingface tiktoken python-bindings mit-license hacker-news pretraining
Read more →

July 15, 2026

MCP 2026-07-28 Goes Stateless: Why the 'USB-C for AI' Is Finally Ready to Sit Behind a Load Balancer

The release candidate for MCP's 2026-07-28 specification — locked May 21, 2026, final-published July 28 — removes the stateful handshake that pinned...

mcp model-context-protocol stateless load-balancer api-gateway enterprise-authorization id-jag oauth open-telemetry w3c-trace-context cache-control anthropic microsoft okta asana atlassian figma slack spring-ai uberconf platform-engineering
Read more →

Archive

Loading...